The way I make games is that first I make a prototype (this is prototyprally after all). Using that I try to find an amusing mechanic, something that makes you smile when you play with it. Normally this isn’t actually all that hard, especially when you’re doing something with physics as it tends to be reasonably emergent. With some fiddling you can find a few interesting mechanics without too much hassle.
The hard part is making it more than a toy.

This is where I’m at with this game now. It’s pretty fun as a toy, and it’s great fun to spawn in a hundred boats and just smash them, but it’s not quite a game yet. But it’ll get there.

My biggest problem right now is resisting the urge to add in even more complicated stuff, blooming and motion blur are two things I’m pretty confident I could put in there without completely killing performance. I’ve also had some problems with my constraint system (used for ropes and such) so that would need a slight rewrite too. But if I’m to complete this within the coming weeks I can’t do that stuff, I need to focus on the actual game.

And once the game’s done there’s menus, possibly high scores and that whole thing. And that takes alot of time, I learned that the hard way for Eater of Worlds, it was supposed to take five weeks, but the highscores and challenge system took two whole weeks on their own pushing the development time to seven weeks total.

Well. Enough with the whining. FlashDevelop, which I plugged in my last post is truly awesome. I hope to be able do a quick little tutorial for how to setup the whole thing for game development.